It is known that only a minority of patients with faecal incontinence report these symptoms to their physicians. Epidemiological estimates based on medical chart data, therefore, may contain a detection bias. To evaluate such bias in epidemiological data, we evaluated prospectively the presence of faecal incontinence in selected patient groups with a proven high incidence of such symptoms and compared it to the incidence in a group of healthy controls. If the patient acknowledged faecal incontinence in the questionnaire, the medical chart was checked to see if these symptoms had been noted during previous work-up; this was used to estimate the number of unregistered cases if the epidemiological estimate is based on medical chart data. The incidence of incontinence was significantly elevated in all patient groups as compared to the controls, but only up to 5% of patients with faecal incontinence, regardless of the underlying mechanism, had these incontinence symptoms noted in the medical charts. We conclude that for the estimation of the prevalence and incidence of faecal incontinence, data from medical charts contain a detection bias which systematically underestimates the real presence of faecal incontinence.